MUC1 is a large, transmembrane glycoprotein expressed on the apical surface of many types of polarized epithelia including pancreas, lung, breast and the gastrointestinal tract (Finn, et al. (1995) Immunol. Rev. 145:61). MUC1 is overexpressed and differentially glycosylated by a number of adenocarcinomas (Croce, et al. (1997) Anticancer Res. 17:4287) and has been evaluated as a candidate antigen for active immunotherapy protocols. Humoral and cell-mediated immune responses against MUC1 are detected in patients with MUC1+ tumors, as measured in vitro (Domenech, et al. (1995) J. Immunol. 155:4766; Petrarca, et al. (1999) Cancer Immunol. Immunother. 47:272; Nakamura, et al. (1998) J. Gastroenterol. 33:354); however, these responses are ineffective at eliminating the tumors in vivo.
A number of MUC1-based immunogens have been evaluated as potential cancer vaccines (Graham, et al. (1996) Int. J. Cancer 65:664; Chien-Hung and Wu (1998) J. Biomed. Sci. 5:231; Reddish, et al. (1998) Int. J. Cancer 76:817; Heukamp, et al. (2002) J. Immunother. 25:46). These include whole cells expressing MUC1, MUC1 purified from tumor cells, and peptide or glycopeptide fragments derived from the tandem repeat region of MUC1 (Finn, et al. (1995) supra; Graham, et al. (1996) Cancer Immunol. Immunother. 42:71; U.S. Pat. Nos. 5,744,144, 5,827,666 WO 88/05054, U.S. Pat. Nos. 4,963,484 and 6,344,203). Clinical trials that utilized MUC1 as a vaccine component focused on the tandem repeat region (Finn, et al. (1995) supra; Graham, et al. (1996) supra; Chien-Hung and Wu (1998) supra; Reddish, et al. (1998) supra). Putative epitopes from regions outside of the tandem repeat region of MUC1 have also been investigated (Brossart, et al. (2000) Blood 96:3102; Brossart, et al. (1999) Blood 93:4309; Heukamp, et al. (2001) Int. J. Cancer 91:385); however, other potentially important epitopes from this tumor-associated antigen, especially those in the cytoplasmic tail, have not been studied. Most studies have used in vitro assays to investigate that the tandem repeat region contains immunodominant epitopes for production of MUC1 specific antibodies and cytotoxic T-lymphocytes (CTL). However, it has been shown that in vitro assays of cytolytic responses do not accurately predict MUC1-specific tumor rejection (Tempero, et al. (1998) J. Immunol. 161:5500). For example, no detectable differences were observed in the anti-MUC1 CTL precursor frequencies of wild-type C57BL/6 mice and C57BL/6 mice transgenic for human MUC1 (MUC1.Tg) (Tempero, et al. (1998) supra), although wild-type mice rejected MUC1-expressing tumors in a MUC1-specific manner while MUC1.Tg mice did not reject these tumors and showed evidence of immunological tolerance to MUC1 (Tempero, et al. (1998) supra; Rowse, et al. (1998) Cancer Res. 58:315).
In vivo immune responses directed against tumor-associated MUC1 have also been analyzed. The nature of cellular immune responses that mediate rejection of MUC1-expressing tumors in mice was investigated by experiments that depleted CD4+, CD8+ or both T cell subsets in vivo. CD4+ cells were required for elimination of a human MUC1-expressing murine melanoma cell line (B16.MUC1), and CD8+ cells were required for the elimination of a human MUC1-expressing murine pancreatic carcinoma cell line (Panc02.MUC1), in wild-type C57BL/6 mice (Tempero, et al. (1999) Int. J. Cancer 80:595; Morikane, et al. (2001) Int. Immunol. 13:233). Studies using mice deficient in molecular components critical to the immune responses (VanLith, et al. (2002) Int. Immunol. 14:873; Sivinski, et al. (2002) Cancer Immunol. Immunother. 51:327) further showed that both CD4+ and CD8+ responses were mediated by α/β T cell receptors and required costimulation through CD28, as well as interactions between CD40 and CD40 ligand, and the activities of interferon γ(IFNγ), and lymphotoxin α. A number of other factors (IL4, IL10, IL12, TNFR-1) were not required. There were differences in the effector mechanisms as the CD8-mediated cytotoxicity required perforin but not FasL; in contrast, the CD4-mediated cytotoxic response required both perforin and FasL.